diff --git a/docs/troubleshooting.md b/docs/troubleshooting.md
index a5502cb0186de3a6034b861ffdbc514895fe6de5..697c14c6788ccf8a44412882b21c63a122310137 100644
--- a/docs/troubleshooting.md
+++ b/docs/troubleshooting.md
@@ -259,8 +259,8 @@ Lets take it as an example how to debug the root cause.
 
 If ever things fail beyond possible recovery, here's how to completely purge an OAS installation in order to start from scratch:
 
-    cluster$ systemctl stop k3s
+    cluster$ /usr/local/bin/k3s-killall.sh
     cluster$ systemctl disable k3s
-    cluster$ mount | egrep '(kubelet|nsfs|k3s)' | cut -d' ' -f 3 | xargs -n 1 umount
+    cluster$ mount | egrep '(kubelet|nsfs|k3s)' | cut -d' ' -f 3 | xargs --no-run-if-empty -n 1 umount
     cluster$ rm -rf /var/lib/{rancher,OpenAppStack,kubelet,cni,docker,etcd} /etc/{kubernetes,rancher} /var/log/{OpenAppStack,containers,pods} /tmp/k3s /etc/systemd/system/k3s.service
     cluster$ systemctl reboot